监视gke上的CPU / mem使用情况

Question

I recently launched with gke and kubernetes in production. I have regular outages with no obvious reasons. No event shows anything, pods are not restarting and seems stable. I have a similar qa env that has no issue at all whereas it's way smaller.

Where can I find potential infos on the outage reason?

Answer 1

You can see monitoring data for your cluster using Stackdriver . There's a brief walkthrough of how to use it for GKE in this blog post . You may also want to check out the general Kubernetes application troubleshooting guide .

What are the symptoms of the outage?

Answer 2

Stack driver makes you pay and configure it... kubernetes comes with a tool for it... just use this:

kubectl top nodes

al@host:~/$ kubectl top nodes
NAME                             CPU(cores)   CPU%      MEMORY(bytes)   MEMORY%
gke-learn-pool-1-10f60e0a-s44c   104m         11%       1008Mi          86%

You can also go under clusters -> Cluster -> nodes -> Node

Update: Stack Driver deprecated all load monitoring plugins. It's K8s or the highway now.